HyperAI超神经

首页资讯最新论文教程数据集百科 SOTA LLM 模型天梯 GPU 天梯顶会

中文

HyperAI超神经

3D Lane Detection On Apollo Synthetic 3D Lane

评估指标

F1

X error far

X error near

Z error far

Z error near

评测结果

各个模型在此基准测试上的表现结果

模型名称	F1	X error far	X error near	Z error far	Z error near	Paper Title	Repository
PersFormer	92.9	0.356	0.054	0.234	0.010	PersFormer: 3D Lane Detection via Perspective Transformer and the OpenLane Benchmark
Gen-LaneNet	88.1	0.496	0.061	0.214	0.012	Gen-LaneNet: A Generalized and Scalable Approach for 3D Lane Detection
Anchor3DLane	95.6	0.306	0.052	0.223	0.015	Anchor3DLane: Learning to Regress 3D Anchors for Monocular 3D Lane Detection
Anchor3DLane† (iterative regression)	95.4	0.299	0.048	0.220	0.013	Anchor3DLane: Learning to Regress 3D Anchors for Monocular 3D Lane Detection
Reconstruct from Top View	91.9	0.387	0.049	0.213	0.008	Reconstruct from Top View: A 3D Lane Detection Approach based on Geometry Structure Prior	-
CurveFormer	95.8	0.326	0.078	0.219	0.018	CurveFormer: 3D Lane Detection by Curve Propagation with Curve Queries and Attention	-
LaneCPP	97.4	0.277	0.030	0.206	0.011	LaneCPP: Continuous 3D Lane Detection using Physical Priors	-
BEV-LaneDet	96.9	0.242	0.016	0.216	0.02	BEV-LaneDet: a Simple and Effective 3D Lane Detection Baseline	-
CLGo	91.9	0.361	0.061	0.25	0.029	Learning to Predict 3D Lane Shape and Camera Pose from a Single Image via Geometry Constraints
3D-LaneNet	86.4	0.477	0.068	0.202	0.015	3D-LaneNet: End-to-End 3D Multiple Lane Detection

0 of 10 row(s) selected.